M2(m-dobdc) (M = Mg, Mn, Fe, Co, Ni) metal-organic frameworks exhibiting increased charge density and enhanced H2 binding at the open metal sites

Article abstract of DOI:10.1021/ja506230r

The well-known frameworks of the type M₂(dobdc) (dobdc ^4- = 2,5-dioxido-1,4-benzenedicarboxylate) have numerous potential applications in gas storage and separations, owing to their exceptionally high concentration of coordinatively unsaturated metal surface sites, which can interact strongly with small gas molecules such as H₂. Employing a related meta-functionalized linker that is readily obtained from resorcinol, we now report a family of structural isomers of this framework, M ₂(m-dobdc) (M = Mg, Mn, Fe, Co, Ni; m-dobdc^4- = 4,6-dioxido-1,3-benzenedicarboxylate), featuring exposed M²⁺ cation sites with a higher apparent charge density. The regioisomeric linker alters the symmetry of the ligand field at the metal sites, leading to increases of 0.4-1.5 kJ/mol in the H₂ binding enthalpies relative to M ₂(dobdc). A variety of techniques, including powder X-ray and neutron diffraction, inelastic neutron scattering, infrared spectroscopy, and first-principles electronic structure calculations, are applied in elucidating how these subtle structural and electronic differences give rise to such increases. Importantly, similar enhancements can be anticipated for the gas storage and separation properties of this new family of robust and potentially inexpensive metal-organic frameworks.

Full text of DOI:10.1021/ja506230r

Article
pubs.acs.org/JACS
M2(m‑dobdc) (M = Mg, Mn, Fe, Co, Ni) Metal−Organic Frameworks
Exhibiting Increased Charge Density and Enhanced H2 Binding at the
Open Metal Sites
Matthew T. Kapelewski,† Stephen J. Geier,†,¶ Matthew R. Hudson,‡ David Stuck,† Jarad A. Mason,†
̈
Jocienne N. Nelson,§ Dianne J. Xiao,† Zeric Hulvey,‡,∥ Elizabeth Gilmour,§ Stephen A. FitzGerald,§
Martin Head-Gordon,†,⊥ Craig M. Brown,‡,# and Jeffrey R. Long*,†,∇
†Department of Chemistry, University of California, Berkeley, Berkeley, California 94720, United States
‡Center for Neutron Research, National Institute of Standards and Technology, Gaithersburg, Maryland 20899, United States
§Department of Physics, Oberlin College, Oberlin, Ohio 44074, United States
∥Department of Materials Science and Engineering, University of Maryland, College Park, Maryland 20742, United States
⊥Chemical Sciences Division and ∇Materials Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720,
United States
#Chemical and Biomolecular Engineering, University of Delaware, Newark, Delaware 19716, United States
S
* Supporting Information
ABSTRACT: The well-known frameworks of the type M2(dobdc) (dobdc4− = 2,5-dioxido-
1,4-benzenedicarboxylate) have numerous potential applications in gas storage and
separations, owing to their exceptionally high concentration of coordinatively unsaturated
metal surface sites, which can interact strongly with small gas molecules such as H2.
Employing a related meta-functionalized linker that is readily obtained from resorcinol, we
now report a family of structural isomers of this framework, M2(m-dobdc) (M = Mg, Mn,
Fe, Co, Ni; m-dobdc4− = 4,6-dioxido-1,3-benzenedicarboxylate), featuring exposed M2+
cation sites with a higher apparent charge density. The regioisomeric linker alters the
symmetry of the ligand field at the metal sites, leading to increases of 0.4−1.5 kJ/mol in the
H2 binding enthalpies relative to M2(dobdc). A variety of techniques, including powder X-
ray and neutron diffraction, inelastic neutron scattering, infrared spectroscopy, and first-
principles electronic structure calculations, are applied in elucidating how these subtle
structural and electronic differences give rise to such increases. Importantly, similar
enhancements can be anticipated for the gas storage and separation properties of this new family of robust and potentially
inexpensive metal−organic frameworks.
polarizing and have strong interactions with many small gas
molecules.3a,4 For example, the well-established M2(dobdc) (M
= Mg, Mn, Fe, Co, Ni, Cu, Zn; dobdc4− = 2,5-dioxido-1,4-
benzenedicarboxylate) structure type, also known as M-MOF-
74, CPO-27-M, or M2(dhtp),5 contains a high density of
exposed metal cations and is especially promising for the
storage of a wide variety of gases. As a result of its compact,
highly charged tetraanionic linker, this framework is endowed
with an exceptionally high density of open metal coordination
sites, which are the primary binding sites for small gas
molecules such as H2, CH4, C2H4, and CO2. Consequently,
M2(dobdc) frameworks have been investigated for numerous
potential applications involving gas adsorption, including H2
storage,3a,4a,c,6 methane storage,4b,7 CO2 capture,8 and hydro-
carbon separations.1i,9 All of these applications take advantage
INTRODUCTION
■
Metal−organic frameworks are a well-known class of porous
materials comprised of inorganic units bridged by coordinating
organic linkers. In addition to possessing high internal surface
areas, their physical and chemical properties can be tuned for
specific applications by judicious choice of the metal center and
organic ligand. These properties lead to applications in storing
and separating gases, in which it is particularly important to
have precise control over the strength and specificity of
interactions between the pore surface and various potential
adsorbates.1 Indeed, strong adsorption sites are often installed
on the pore surface in order to attract specific gas molecules
selectively for separation applications or to increase the density
of gas molecules present for storage applications.1h,j,2
Exposed metal cations represent an important example of
strong adsorption sites that have been realized in many metal−
organic frameworks.3 These Lewis acidic sites, which are
typically formed by removing metal-coordinated solvent
molecules upon heating under reduced pressure, are highly
Received: June 20, 2014
Published: August 18, 2014
© 2014 American Chemical Society
12119
dx.doi.org/10.1021/ja506230r | J. Am. Chem. Soc. 2014, 136, 12119−12129

Journal of the American Chemical Society
Article
of strong interactions between gas molecules and exposed
metal cations.
scattering, infrared spectroscopy, and first-principles electronic
structure calculations, we provide a careful comparison of the
differences between M2(dobdc) and M2(m-dobdc) that
contribute to the differences in their H2 adsorption properties.
Discovering new frameworks with a high density of open
metal coordination sites is not trivial, as it is difficult to predict
what topologies will form with partial solvation of the metal
nodes and will then further be amenable to desolvation. Rather,
selectively tuning the most promising of the numerous existing
frameworks by altering the linker, the metal, or both provides a
viable strategy for developing promising new adsorbents. Since
the M2(dobdc) series of frameworks has high chemical and
thermal stability8b,10 and has been shown to be outstanding for
binding a high capacity of small gas molecules, we hoped to
modify this framework by tuning the electronics of the exposed
metal cations, which should subsequently tune the affinity for
different gas molecules. For example, the binding of H2 or CH4
might be enhanced by increasing the positive charge density at
the metal cation site, thus increasing its ability to polarize and
bind adsorbing gas molecules more strongly. Furthermore,
M2(dobdc) offers the advantage of being an isostructural series
of frameworks that can be formed with a variety of metals,
thereby offering an additional level of control for tuning the
framework for specific interactions. Based on the high density
of exposed metal cations, isostructural nature of the these
frameworks, and thermal stability, a combination of properties
that is not paralleled by any other structure type, M2(dobdc)
provides an ideal platform for exploring the tunability of
metal−organic frameworks to strengthen framework−gas
interactions.
This work outlines the synthesis of a structural isomer of the
M2(dobdc) framework, which will be referred to as M2(m-
dobdc) (m-dobdc4− = 4,6-dioxido-1,3-benzenedicarboxylate).
Rather than having para carboxylic acid functionalities and para
hydroxyl substituents, as in the regular H4dobdc linker, H4(m-
dobdc) has meta carboxylic acid groups and meta hydroxyl
substituents. The H4(m-dobdc) linker was targeted due to its
ease of synthesis and potentially low cost. Indeed, a solvent-free
Kolbe−Schmitt reaction to produce H4(m-dobdc) using only
resorcinol, KHCO3, and CO2 is efficient and inexpensive. In
spite of this linker isomerism, it is still possible to form a
framework with a similar overall structure, replete with one-
dimensional hexagonal channels and a high density of open
metal coordination sites. This structural isomerism takes
advantage of the fact that M2(dobdc) frameworks have multiple
types of coordinating functional groups in the linker, which is a
rarity among most well-known metal−organic frameworks.
The resulting new framework has subtle differences in the
electronic structure of the ligand and overall connectivity as
compared to the M2(dobdc) framework, which might be
expected to influence interactions with small gas molecules by
changing the local environment around the open metal
coordination sites. Isomers of metal−organic frameworks are
known and primarily derive from what are termed “framework
isomers.” The structure of M2(m-dobdc), however, is an
example of a ligand-originated isomer.11 Relatedly, a recent
M2(dobdc) analogue was synthesized with thiols rather than
phenols,12 but a purely structural isomer of M2(dobdc) has
never before been synthesized.
EXPERIMENTAL SECTION
■
Methanol was purchased from commercial vendors, further dried over
molecular sieves, and deoxygenated by purging with N2. All other
reagents were purchased from commercial vendors and used without
further purification, unless otherwise noted. The H4(m-dobdc) ligand
was synthesized according to a literature procedure.14
Synthesis of M2(m-dobdc) (M = Mn, Fe, Co, Ni). Anhydrous
MCl2 (3.0 mmol) and H4(m-dobdc) (240 mg, 1.2 mmol) were added
to 80 mL of a mixed solvent (x% MeOH by volume in
dimethylformamide (DMF), where for Mn, x = 15; Fe, x = 15; Co,
x = 50; Ni, x = 35) in a nitrogen-filled glovebox for M = Mn, Fe and in
air for M = Co, Ni. The solution was dispensed into eight 20 mL
scintillation vials, which were each sealed with a PTFE-lined cap and
heated at 120 °C for 18 h. The resulting solid from each vial was
combined, submerged in 20 mL of DMF, and heated at 70 °C for 24 h.
The DMF was decanted and replaced with 20 mL of methanol. The
resulting suspension was heated at 70 °C for 4 days, during which time
the methanol was replaced every 24 h. The material was activated by
heating it at 150 °C under dynamic vacuum on a Schlenk line for 12 h,
followed by further activation of a small amount of the sample by
heating the solid under dynamic vacuum (<10 μbar) at 180 °C (160
°C for Fe2(m-dobdc)) for 24 h at a ramp rate of 0.5 °C/min. It should
be noted that larger scale syntheses of Co2(m-dobdc) and Ni2(m-
dobdc) were also accomplished with no loss in crystallinity or surface
area by stirring at room temperature in a round-bottom flask equipped
with a reflux condenser at a concentration of 20.2 mmol H4(m-dobdc)
and 50.5 mmol MCl2 (M = Co, Ni) in 1250 mL solvent. Times and
temperatures were identical to the small-scale synthesis, but 200 mL of
solvent was used for each exchange.
Synthesis of Mg2(m-dobdc). This compound was synthesized in
air in a round-bottom flask by adding 300 mg of Mg(NO3)2·6H2O and
93 mg of H4(m-dobdc) to 14 mL MeOH in 31 mL DMF and stirring
at 120 °C for 8 h, then filtering off the resulting solid. Detailed
characterization of all frameworks is contained in the Supporting
Information.
Physical Measurements. Thermogravimetric analyses were
carried out at a ramp rate of 2 °C/min under a 25 mL/min N2 flow
with a TA Instruments TGA Q5000. Infrared spectra were collected
on a PerkinElmer Avance Spectrum 400 FTIR spectrometer equipped
with a Pike attenuated total reflectance (ATR) accessory. Diffraction
data were collected with 0.02° steps using a Bruker AXS D8 Advance
diffractometer equipped with Cu−Kα radiation (λ = 1.5418 Å), a
Gobel mirror, a Lynxeye linear position-sensitive director, and
̈
mounting the following optics: fixed divergence slit (0.6 mm),
receiving slit (3 mm), and secondary beam Soller slits (2.5°). The
generator was set at 40 kV and 40 mA. Samples were either loaded on
zero background sample holders or packed into air-free capillaries in a
nitrogen-filled glovebox and mounted using a capillary stage.
Elemental analyses were obtained from the Microanalytical Laboratory
of the University of California, Berkeley.
Gas Adsorption Measurements. Gas adsorption isotherms for
pressures in the range 0−1.2 bar were measured using a volumetric
method using either a Micromeritics ASAP2020 or ASAP2420
instrument. Samples were transferred under a dinitrogen atmosphere
to preweighed analysis tubes, then capped with a Transeal. The
samples were evacuated at elevated temperature until the outgas rate
was <1 μbar/min, at which point the tube was weighed to determine
the mass of the activated sample, which was typically 50−200 mg. The
tube was transferred to the analysis port of the instrument and the
outgas rate was again checked to ensure that it was <1 μbar/min.
UHP-grade (99.999% purity) N2, H2, and He were used for all
adsorption measurements. For all isotherms, warm and cold free
spaces were measured using He; N2 and H2 isotherms at 77 and 87 K
were measured in liquid nitrogen and liquid argon baths, respectively.
Here, H2 adsorption is used as a probe to determine whether
this new framework exhibits enhanced gas adsorption proper-
ties. The choice of H2 as a probe molecule originates from its
simplicity, low polarizability, and potential use as a clean,
renewable fuel.2a,13 Through the use of H2 adsorption
isotherms, powder neutron diffraction, inelastic neutron
12120
dx.doi.org/10.1021/ja506230r | J. Am. Chem. Soc. 2014, 136, 12119−12129

Journal of the American Chemical Society
Article
Oil-free vacuum pumps and oil-free pressure regulators were used for
all measurements. Brunauer−Emmett−Teller (BET) and Langmuir
surface areas were determined from N2 adsorption data at 77 K using
Micromeritics software.
DFT Calculations. Due to the extended nature of the M2(dobdc)
and M2(m-dobdc) structures, cluster modeling was completed on the
linker of interest coordinated to a pair of Co atoms bound to either
end of the organic linker. To truncate the system, the ligating oxygen
atoms that are not part of the included linker were added as
formaldehyde molecules in order to conserve charge. The
experimentally determined crystal structures were truncated as
described and frozen. The geometry of a hydrogen molecule bound
to the frozen system based on neutron diffraction data was then
optimized. The range-separated, dispersion corrected functional
ωB97X-D implemented in the electronic structure software Q-
Chem18 was used with an ultra fine (99, 590) grid and a triple split-
valence basis set with polarization (6-311G**).19 A small core
Stuttgart−Born (SRSC) effective core potential is employed to model
the core electrons of the Co.20 Binding is further analyzed using the
ALMO EDA.21 Charge transfer is accounted for using the perturbative
Roothaan step approach, which allows for assignment of forward and
backbonding energies as well as generation of complementary
occupied-virtual orbital pairs (COVPs)22 to visualize charge transfer.
Powder Neutron and X-ray Diffraction Experiments. Neutron
powder diffraction (NPD) experiments were carried out on 0.8358,
0.9567, and 0.9702 g activated Co2(m-dobdc), Ni2(m-dobdc), and
Co2(dobdc) samples respectively, using the high-resolution neutron
powder diffractometer, BT1, at the National Institute of Standards and
Technology Center for Neutron Research (NCNR). The samples
were placed in a He purged glovebox and loaded into a vanadium
sample can equipped with a valve for gas loading, and sealed using an
indium O-ring. NPD data were collected using a Ge(311)
monochromator with an in-pile 60′ collimator corresponding to a
wavelength of 2.078 Å. The samples were loaded onto bottom-loading
closed cycle refrigerators, and initial data collected on the activated
Co2(m-dobdc) and Ni2(m-dobdc) frameworks at 10 K. As part of the
initial structure solution, X-ray diffraction (XRD) measurements were
carried out on 12.8 mg of Co2(m-dobdc) at the Advanced Photon
Source (APS) on the 17-BM materials diffractometer (λ = 0.7291 Å)
at 298 K. The activated Co2(m-dobdc) sample was transferred into
quartz capillary in a He purged glovebox and wax sealed for the X-ray
measurements. For comparison of the D2 structural dependence on
ligand connectivity, Co2(dobdc) and Co2(m-dobdc) were each
individually connected to a gas manifold of known volume and
exposed to a known dose, approximately 0.75 and 2.25 D2 per Co2+, at
100 K (refined composition given in Tables S12−S14, S19). Both
samples were slow cooled from 100 to 10 K to ensure full equilibration
and complete adsorption, as evidenced by a zero pressure reading on
the barometer by 25 K, for data collection. Further D2 structural data
were collected on Co2(m-dobdc) and Ni2(m-dobdc) as a function of
dose with final loadings of 0.75, 1.25, 1.5, 1.75, 2.0, 2.25, and 3.0 D2
per Co2+ and 1.0, 2.0, and 3.0 D2 per Ni2+. Structural data for H2 were
measured on Co2(m-dobdc) at a loading of 0.7 H2 per Co2+. Refined
compositions are given in Tables S12 and S15−S20 (Co) and Tables
S21−S23 (Ni). Details on the structure solutions are available in the
Supporting Information.
RESULTS AND DISCUSSION
■
Structural Characterization of M2(m-dobdc). A less
expensive regioisomer of H4dobdc was selected in order to
form a framework with the same overall topology and a high
density of open sites as in M2(dobdc), but with potentially
different local geometry and electronic properties. After the
synthesis of H4(m-dobdc) from resorcinol, it was possible to
synthesize an isostructural series of M2(m-dobdc) (M = Mg,
Mn, Fe, Co, or Ni) frameworks (Figure 1) exhibiting powder
Inelastic Neutron Scattering. Inelastic neutron scattering (INS)
spectra were collected using the Filter Analyzer Neutron Spectrometer
(FANS)15 at the NCNR on the same samples used for the NPD
experiments. Spectra were obtained at 7 K using the pyrolytic graphite
(002) monochromator and 20′-20′ collimation options. Data were first
collected for the bare framework, followed by data collection for the
sample loaded with normal-H2 (n-H2), which contains a 3:1 mixture of
ortho (o-H2) to para (p-H2), respectively. For Co2(m-dobdc), data
were collected at loadings of 0.33, 0.5, 1.0, 2.0, 3.0, and 4.0 n-H2
molecules per Co atom, while for Ni2(m-dobdc) data were collected at
loadings of 0.67, 1.0, 2.0, 3.0, and 4.0 n-H2 molecules per Ni atom. Gas
was loaded into the materials using the same methodology as
described in the NPD experiments. The spectra of the bare
frameworks were subtracted from the spectra obtained from the H2
loaded samples, and Gaussian peaks were fit to the rotational
transitions using the DAVE suite of programs.16 Further measure-
ments of the framework vibrational densities of states for the activated
Co2(m-dobdc) material were made to higher energies (35 to 160
meV) using the Cu(220) monochromator with 20′-20′ collimation.
Infrared Spectroscopy. Infrared spectra were acquired using a
Bomem DA3Michelson interferometer equipped with a glowbar
source, CaF2 beamsplitter, and a liquid nitrogen cooled mercury−
cadmium-telluride detector. A cutoff filter above 9000 cm−1 was used
to prevent unwanted sample heating from the IR source. A custom-
built diffuse reflectance system17 with a sample chamber that allows
both the temperature and atmosphere of the material to be controlled
was used for all experiments. Powder samples of the frameworks (∼10
mg) were transferred under inert atmosphere to a cup affixed to a
copper slab providing thermal contact to a coldfinger cryostat (Janis
ST-300T). The sample temperature was monitored by a Si-diode
thermometer. Known quantities of H2 gas were dispensed from a
calibrated gas manifold by monitoring the change in pressure.
Figure 1. Synthesis of H4(m-dobdc) and M2(m-dobdc), starting from
the inexpensive compound resorcinol.
X-ray diffraction patterns (Figures S1−S5) analogous to the
respective M2(dobdc) frameworks, indicating that the two
series adopt similar structures. Thermogravimetric analyses of
the Co2(m-dobdc) and Ni2(m-dobdc) frameworks show initial
mass losses of 33% and 14%, respectively, indicative of
volatilization of trapped solvent molecules and thus porosity
(Figure S6a). After repeatedly washing each compound with
methanol to replace the metal-bound DMF, the frameworks
were activated by heating under dynamic vacuum at 180 °C.
Thermogravimetric analysis of the activated frameworks
indicated that Mn2(m-dobdc) and Co2(m-dobdc) are stable
to 250 °C and the Ni2(m-dobdc) is stable to 350 °C.
To confirm that the M2(m-dobdc) frameworks are indeed
structurally analogous to M2(dobdc), powder X-ray and
neutron diffraction experiments were used to solve the crystal
structure of Co2(m-dobdc) (Figure 2). Similar to Co2(dobdc),
Co2(m-dobdc) possesses helical chains of Co2+ centers running
parallel to the crystallographic c axis. Upon activation and
removal of a bound DMF molecule, each metal center is ligated
by a combination of oxido and carboxylate donors in a square
pyramidal geometry, with open coordination sites directed into
the one-dimensional hexagonal channels of the framework.
Based on the change in point group symmetry of the linker
12121
dx.doi.org/10.1021/ja506230r | J. Am. Chem. Soc. 2014, 136, 12119−12129

Journal of the American Chemical Society
Article
has proven challenging, and all efforts have thus far resulted in
materials with lower than expected surface areas. As a result, the
remaining data will focus only on the other four M2(m-dobdc)
compounds.
H2 Adsorption Isotherms. To probe the potentially
modified electronic structures at the open metal coordination
sites, low-pressure H2 adsorption isotherms were measured for
the four M2(m-dobdc) (M = Mn, Fe, Co, Ni) frameworks at 77
and 87 K (Figures S20−S23). As for the M2(dobdc) series,4a
Ni2(m-dobdc) has the greatest H2 uptake at 77 K and 1 bar,
followed by Fe, Mn, and Co compounds, respectively. Although
the M2(dobdc) frameworks have a higher H2 uptake at 1 bar, in
the very low-pressure regime, Ni2(m-dobdc) has a significantly
higher uptake than Ni2(dobdc) at both 77 and 87 K (Figure
S27). This indicates that the interaction of H2 with the exposed
Ni2+ cations ion the surface of Ni2(m-dobdc) is stronger than in
Ni2(dobdc); since these frameworks have about the same
surface area, another effect must be influencing the binding
strength of H2.
Figure 2. Crystal structure of Co2(m-dobdc) showing one-dimensional
Isosteric heats of H2 adsorption were calculated in order to
gain insight into this increased low-pressure H2 adsorption and
confirm whether there is indeed a stronger interaction between
H2 and the metal centers in the M2(m-dobdc) frameworks. In
order to determine the isosteric heat of adsorption (Qst) using
the Clausius−Clapeyron relation, H2 isotherm data at 77 and
87 K were fit with either a dual-site Langmuir−Freundlich
equation or a trisite Langmuir equation. Similar to
M2(dobdc),4a the isosteric heat of adsorption plots for M2(m-
dobdc) imply a nearly constant H2 binding enthalpy until a
loading of 0.7−0.8 H2/M2+, followed by a sharp decrease as all
exposed metal cations become occupied and only weaker
adsorption sites are available (Figure 3). The inflection points
hexagonal pores and helical metal chains.
from C2h in H4(dobdc) to C2v in H4(m-dobdc), a change in the
space group from R3 in Co (dobdc) to R3m in Co (m-dobdc)
̅
2
2
is observed.4a Additional structural differences between the two
compounds are apparent.9e For instance, the orientation of the
carboxylate groups of the linker is changed, with the CO2− unit
twisting out of the plane of the aromatic ring by approximately
12.5° in Co2(m-dobdc), as compared to just 3.5° in
Co2(dobdc). Furthermore, metal centers in Co2(m-dobdc)
that face into the same pore align directly along the
crystallographic b axis, resulting in a Co···Co separation of
14.9(1) Å across the channels. In Co2(dobdc), the metals are
offset from each other by one-third of a twist in the chain,
leading to a Co···Co separation of 15.24(8) Å across the
channels. A structureless Le Bail refinement of the other M2(m-
dobdc) (M = Mg, Mn, Fe, Ni) compounds afforded related unit
cells, confirming that the entire series is isostructural.
With such similar crystal structures, the surface areas would
be expected to be similar for the two types of frameworks.
Indeed, a geometric approximation of the surface areas for
Co2(m-dobdc) and Co2(dobdc) predicts them to both be 1297
m2/g.23 Low-pressure N2 adsorption isotherms collected at 77
K reveal type I isotherms characteristic of microporous solids
for the M2(m-dobdc) (M = Mn, Fe, Co, Ni) compounds
(Figure S11). Langmuir and BET surface areas were calculated
(Table 1), and in each case the results are comparable to those
reported for M2(dobdc).4b,c,9e,24 Based on the similar frame-
work structure, this indicates full evacuation of the pores and
complete desolvation of the framework. Furthermore, the
Co2(m-dobdc) BET surface area of 1264 m2/g is very close to
the predicted surface area of 1297 m2/g from the geometric
calculations. Achieving complete activation of the Mg analogue
Figure 3. H2 isosteric heat of adsorption curves for the M2(m-dobdc)
series of frameworks, as a function of the amount adsorbed.
range from ∼0.75 H2/M in Co2(m-dobdc) to ∼0.85 H2/M in
Fe2(m-dobdc), which corresponds with the fraction of metal
sites available for H2 binding. These values are comparable to
previously observed values for the M2(dobdc) series.24
Increasing the binding enthalpy of H2 in metal−organic
frameworks is important for practical applications, as a binding
enthalpy of −15 to −20 kJ/mol is predicted to be optimal for
the on-board storage of H2 at ambient temperatures.25 The
low-coverage isosteric heats of adsorption of Mn2(m-dobdc),
Fe2(m-dobdc), Co2(m-dobdc), and Ni2(m-dobdc) are −10.3,
−11.1, −11.6, and −12.3 kJ/mol, respectively. This trend in
binding enthalpies for H2 is consistent with the Irving−
Table 1. Langmuir and BET Surface Areas of the M2(dobdc)
a
and M2(m-dobdc) Frameworks
Mn4b,9e
Fe4c
Co24a
Ni24b
1574
Langmuir (m2/g)
BET (m2/g)
1797
1102
1741
1349
1535
1360
1624
1295
1432
1341
1504
1264
M2(dobdc)
Langmuir (m2/g)
BET (m2/g)
1592
1321
M2(m-dobdc)
a
The values for M2(dobdc) are from literature sources.
12122
dx.doi.org/10.1021/ja506230r | J. Am. Chem. Soc. 2014, 136, 12119−12129

Journal of the American Chemical Society
Article
Williams series, which predicts that high-spin complexes
increase in stability moving from Groups 7 to 10.26 The
trend further mirrors that observed for H2 binding within
M2(dobdc) frameworks (Table 2). Calculated isosteric heat of
(Table 2) and are as much as 1.5 kJ/mol stronger in the case of
Mn, which also has the largest percent increase (17%) in
binding enthalpy.
Powder Neutron Diffraction. In an effort to understand
further why the M2(m-dobdc) frameworks show higher H2
binding enthalpies than the structurally similar M2(dobdc),
neutron diffraction experiments were performed on micro-
crystalline powder samples dosed with precise quantities of D2.
Data were collected upon successively dosing the evacuated
Co2(m-dobdc) and Ni2(m-dobdc) samples with loadings
ranging from 0.75 to 3.0 D2 molecules per metal center and
cooling to 10 K. Based on previous results for M2(dobdc)
frameworks6c as well as the isosteric heats of adsorption for
M2(m-dobdc) that begin to decrease at loadings near 1 H2/
metal, it was anticipated that the open metal coordination site
would provide the primary hydrogen binding site. Indeed, this
site, designated site I, was found to be the only D2 binding site
in Co2(m-dobdc) at the lowest loading of 0.75 D2 per Co2+
(Figure 4a). The center of the D2 electron density was observed
to be at a separation of 2.23(5) Å from the metal center, closer
than the 2.32(2) Å found for Co2(dobdc) at the same loading
(Tables S12−S13), which is further confirmation that the
hydrogen binds more strongly to the open metal sites within
the meta framework. The same binding site is apparent in
Ni2(m-dobdc), with a comparable M···D2 distance of 2.18(4) Å
at a loading of 1.0 D2 per Ni2+, which is within error of the
distance of 2.201(1) Å observed for Ni2(dobdc) at 4 K.27
Increasing the D2 loading in Co2(m-dobdc) revealed several
secondary binding sites. At a loading of 1.25 D2 per metal, a
second binding site (site II) adjacent to the primary site
becomes populated (Figure 4b). This can be attributed to a
D2···D2 interaction based on a short D2···D2 separation of
2.88(4) Å, combined with a D2···O interaction occurring at a
Table 2. Comparison of Low-Loading H2 Isosteric Heats of
a
Adsorption (Qst, kJ/mol) in M2(dobdc) and M2(m-dobdc)
Mn
Fe
Co
Ni
M2(dobdc)
−8.8
−10.3
−9.7
−11.1
−10.8
−11.5
−11.9
−12.3
M2(m-dobdc)
a
The M2(dobdc) values are from this work to make an accurate
comparison except for Fe2(dobdc), which was previously reported by
our group and fit in the same manner.4c Other literature values are also
available.4a
adsorption values are highly sensitive to the fitting equation
used, the temperatures at which data are collected and the
quality of the fits. Given this, the values for the M2(dobdc)
series were calculated from isotherms collected in this work
(except Fe2(dobdc), which was previously reported by our
group and fit using the same method)4c in order to maintain
consistency in the collected isotherms and the manner in which
they are fit. These values all agree with the literature values for
the M2(dobdc) series except for Ni2(dobdc), which we found
to be 1 kJ/mol less than the literature value of −12.9 kJ/mol.4a
This discrepancy most likely arises from a difference in data
collection, isotherm temperatures, and equation used to fit the
data. Consequently, our data that were collected and fit in the
same manner as the M2(m-dobdc) data presented here are used
for comparison. Importantly, the isosteric heats of adsorption
for the M2(m-dobdc) compounds are, on average, ∼1.0 kJ/mol
stronger than in the corresponding M2(dobdc) frameworks
Figure 4. Partial crystal structures at 10 K of Co2(m-dobdc) showing (a) the primary binding site (site I) on the open metal site; (b) binding site II
interacting with the D2 in site I; (c) binding site III on the aromatic ring in the linker; and (d) binding site IV, interacting with the D2 in site III.
Purple, gray, red, and white spheres represent Co, C, O, and H atoms, respectively. Yellow and orange spheres represent D2 molecules.
12123
dx.doi.org/10.1021/ja506230r | J. Am. Chem. Soc. 2014, 136, 12119−12129

Journal of the American Chemical Society
Article
separation of 3.28(6) Å from the nearest framework oxygen
atom. Site II is in a similar location to the secondary binding
sites observed in Co2(dobdc) and Mg2(dobdc),6c whereby the
second bound D2 appears to rely on D2···D2 interactions in all
cases. The site I to site II D2···D2 separation, however, of
3.05(2) Å in Co2(m-dobdc) at a loading of 2.25 D2 per metal is
significantly shorter than the 3.16(2) Å observed at the same
loading in Co2(dobdc) or the 3.16(8) Å arising for a similar
loading of 1.2 D2 per metal in Mg2(dobdc). This closer site I to
site II separation in Co2(m-dobdc) is most likely a polarization
effect, as the more strongly bound D2 residing at site I has a
larger induced dipole, leading to a stronger interaction with a
D2 molecule adsorbed at site II.
in Figure 5. At loadings up to and including 1.0 n-H2
(corresponding to a 3:1 ortho-H2:para-H2 ratio) molecule per
A third binding site (site III) located approximately 3.08 Å
above the mean plane of the aromatic ring of the m-dobdc4−
linker becomes populated at higher loadings of D2. This site is
equidistant (3.25(6) Å) to two D2 molecules at neighboring site
I positions in Co2(m-dobdc), whereas in Co2(dobdc), the
different symmetry leads to two distinct site I···site III
interactions at distances of 3.06(3) and 3.24(3) Å. Importantly,
for Co2(m-dobdc) there are only half as many D2 molecules
situated at site III as in Co2(dobdc) at high loadings, which
leads to the observation of a fourth binding site (site IV) not
seen in the M2(dobdc) series. We note that a fourth site was
identified in Zn2(dobdc),6b but this is not the same site seen in
the Co2(m-dobdc) case. The differing symmetry of the meta
framework contributes to this additional binding site, as the two
framework oxygen atoms interacting with the D2 molecule in
site IV are equivalent in Co2(m-dobdc), but not in Co2(dobdc).
Relative to site I, larger atomic displacement parameters for the
refined D2 molecules at binding sites II, III, and IV are also
indicative of weaker interactions. Neutron diffraction measure-
ments of Ni2(m-dobdc) with different loadings of D2 resulted in
the identification of the same binding sites as those found for
Co2(m-dobdc) (Tables S21−S23).
Neutron diffraction experiments intended to elucidate
binding sites are generally performed using D2 instead of H2
because the large incoherent scattering cross-section of H
atoms contributes strongly to the background and reduces data
quality. In addition, the coherent scattering cross-section of a D
atom is approximately four times larger than that of an H atom.
While it has been shown through gas adsorption measurements
that metal−organic frameworks tend to adsorb D2 slightly more
strongly than H2,28 it has been assumed that both molecules
behave similarly when binding, only displaying extremely minor
differences in interaction distances. In order to test this
assumption, we measured and successfully refined the structure
of Co2(m-dobdc) dosed with a low loading of H2. We believe
this is the first structural data from powder neutron diffraction
reported for H2 in a metal−organic framework. As expected
from the D2 data, a dose below 1.0 H2 per metal site resulted in
binding exclusively at site I. Interestingly, the center of the H2
density was found to be 2.26(4) Å from the Co atom, perhaps
slightly longer than the Co···D2 distance of 2.23(5) Å, although
still within the associated error. Whether this is indicative of a
slightly stronger interaction with D2 is somewhat unclear due to
the resolution of the data; however, it does confirm that H2 and
D2 can be expected to behave similarly for the purposes of
structural determination.
Figure 5. INS data for Co2(m-dobdc) (left) and Ni2(m-dobdc) (right)
at loadings of 0.33 (black), 0.50 (dark green), 0.67 (light green), 1.0
(red), 2.0 (blue), 3.0 (yellow), and 4.0 (purple) n-H2 molecules per
metal atom. Data are shown after subtraction of the spectrum of the
evacuated framework and offset for clarity.
metal atom, two low-energy rotational lines are apparent at
7.8(1) and 9.5(1) meV for Co2(m-dobdc) and 7.5(1) and
9.3(1) meV for Ni2(m-dobdc). These features are similar to
those observed in INS spectra for several compounds in the
M2(dobdc) series at low loadings and have been assigned to
transitions occurring from the J = 0 state to sublevels of the
split J = 1 rotational state for initial H2 molecules adsorbed at
the metal centers.4c,6b,c,24 This assignment has been confirmed
through correlation with neutron diffraction and DFT
calculations.29 Presumably, there is also a higher-energy
transition, not collected within this current data range, as
seen in the M2(dobdc) series. The splitting between the low-
energy peaks of approximately 1.6 meV for both Co2(m-dobdc)
and Ni2(m-dobdc) is smaller than that observed for any of the
M2(dobdc) materials. The position of the first peak is at higher
energies than that of the first peak of the M2(dobdc)
frameworks (except for the Zn analogue), and the position of
the second peak is at lower energies than is observed for all the
M2(dobdc) compounds. The energy splitting of these peaks has
been previously shown to have no correlation with the binding
strength of H2 at the open metal site.30
At loadings above 0.75 H2 molecules per metal, as sites II−IV
begin to populate, the additional H2 molecules affect the
rotational potential for H2 molecules at site I. This adjusts the
rotational energy level, resulting in a shift of the first peak to
lower energies and the second peak to higher energies. These
peak shifts are similar to those observed for the M2(dobdc)
frameworks, with magnitudes of approximately 0.7 meV for
Co2(m-dobdc) and 0.5 meV for Ni2(m-dobdc), though much
less than the 1.4 meV shift observed for Fe2(dobdc).4c
Adsorption of H2 at the secondary sites also results in a
significant increase in the area of the peak at 9 meV and
appearance of features near 12 and 15 meV as a new subset of
rotational levels associated with the rotationally hindered
second adsorption site. Transitions in this energy range are
also observed in spectra reported for higher loadings of the
M2(dobdc) materials, indicating the similarity in adsorption
potentials at these secondary sites.4c,6c
Infrared Spectra. Since the higher binding enthalpies and
low-coverage H2 uptake cannot reasonably be attributed to a
significant change in the macrostructure of the framework,
infrared spectroscopy was used to further probe the binding of
H2 at the open metal coordination sites. The vibrational
frequency of adsorbed H2 is almost always lower than that of
Inelastic Neutron Scattering. Inelastic neutron scattering
(INS) experiments were carried out to probe the site-specific
binding properties of H2. Data for various loadings of H2 after
subtraction of the spectra for the evacuated materials are shown
12124
dx.doi.org/10.1021/ja506230r | J. Am. Chem. Soc. 2014, 136, 12119−12129

Journal of the American Chemical Society
Article
the molecule in the gas phase (4161 cm−1), and it is now well
established that for metal−organic frameworks, there is a strong
correlation between the magnitude of the frequency shift and
the binding energy at a particular site.31 A comparison of the H2
vibrational frequencies is shown at two different H2 loadings for
Ni2(m-dobdc), Ni2(dobdc), Co2(m-dobdc), and Co2(dobdc)
(Figure 6) and for Mn2(m-dobdc) and Mn2(dobdc) (Figure
Variable-temperature infrared spectroscopy is a standard
technique for establishing the enthalpy of adsorption at a
particular site.32 Figure 7 shows the spectra obtained for H2 in
Figure 7. Variable-temperature infrared spectra of Co2(m-dobdc). The
inset shows the van’t Hoff plot that is used to extract the enthalpy and
entropy change in H2 upon adsorption to the open metal site.
Co2(m-dobdc) while lowering the sample temperature from
142 to 75 K. In each case, the initial H2 pressure for the system
was set such that only the open metal site was occupied over
the full temperature range. The fractional occupancy is then
determined by the ratio of the area under the infrared band to
that observed at complete saturation. The inset in Figure 7
shows the van’t Hoff relationship plot used to extract both the
enthalpy and entropy change upon adsorption. The data
obtained in this way for the different materials are summarized
in Table 3. The slopes extracted from the van’t Hoff plots are
Figure 6. Comparison of the IR spectra of Ni2(m-dobdc), Ni2(dobdc),
Co2(m-dobdc), and Co2(dobdc) at two different concentrations. The
right peak shows the H2 bound to the open metal site, and the left
peak shows the H2 bound to the secondary adsorption sites. Spectra
are offset for clarity.
S45). At this temperature, the pure vibrational part of the
spectrum consists of an ortho−para pair that is separated by just
6 cm−1 in the gas phase. The peak near 4025 cm−1 in each
spectrum corresponds to the H2 bound to the open metal site.
In each case, the spectra for loadings of 0.5 H2 molecules per
metal, which are known to arise purely from H2 bound to the
open metal site, show a greater frequency shift for the Co2(m-
dobdc) and Ni2(m-dobdc) than for Co2(dobdc) or Ni2(dobdc).
This shift is consistent with the stronger H2 binding observed
in the M2(m-dobdc) frameworks based on equilibrium
adsorption isotherms and neutron diffraction studies. Upon
further loading, the secondary adsorption sites begin to be
populated, which is evidenced by a second peak that grows in
around 4125 cm−1. The energy of the peaks associated with the
secondary sites is very similar to those seen for M2(dobdc),
consistent with the INS transitions for H2 bound to the
secondary sites, which were also similar between M2(m-dobdc)
and M2(dobdc) compounds. This supports our hypothesis that
the electronic structure at the metal center is significantly
altered with the m-dobdc4− linker, leading to a higher initial
isosteric heat of adsorption, while the secondary sites in M2(m-
dobdc) and M2(dobdc) are similar in binding potential due to
structural similarities between the two frameworks. Further-
more, a consistent shift in the frequency for the metal-adsorbed
H2 molecules as a function of concentration is seen for both the
Co2(m-dobdc) and Ni2(m-dobdc) samples, although not to the
same degree in the Mn2(m-dobdc) sample. These concen-
tration shifts have previously been attributed to H2···H2
interactions, and it is interesting to note that the shifts seem
largely unaffected by the change in linker. Furthermore, these
shifts correlate well with the shifts seen in rotational potential
as the secondary binding sites are populated in the INS data
shown in Figure 5.
Table 3. Enthalpy and Entropy Changes upon Adsorption of
a
H2 to the Open Metal Site
M2(dobdc)
M2(m-dobdc)
M
ΔS (J/mol K) ΔH (kJ/mol) ΔS (J/mol K) ΔH (kJ/mol)
Mn
Co
Ni


−135
−147
−147
−10.5
−12.1
−13.7
−136
−148
−10.7
−12.3
a
Values are extracted from a van’t Hoff plot based on the area of the
infrared vibrational peaks for a given temperature and pressure.
sensitive to small variations in the maximum saturation area; as
such, an error on the order of 0.5 kJ/mol is estimated. Even
with these limitations, the data is consistent with the calculated
isosteric heats of adsorption from the H2 adsorption isotherms,
showing an enthalpy increase from Mn to Co to Ni. More
importantly, these infrared spectroscopy-based heat of
adsorption values are about 1.4 kJ/mol larger than those
calculated for their respective M2(dobdc) counterparts from
similar infrared experiments. The binding enthalpy of −13.7 kJ/
mol in Ni2(m-dobdc) is among the highest reported H2 binding
enthalpies in a metal−organic framework to date. This is
consistent with the trends observed using values calculated
from H2 adsorption isotherms. We further note that the
measured large entropy changes are consistent with previous
studies, in which a strong correlation between the enthalpy and
entropy change of the bound hydrogen in different metal−
organic frameworks was observed.33 Overall, the redshift seen
12125
dx.doi.org/10.1021/ja506230r | J. Am. Chem. Soc. 2014, 136, 12119−12129

Journal of the American Chemical Society
Article
in the infrared spectra for the adsorbed H2 in M2(m-dobdc) as
compared to M2(dobdc) indicates that the H2 is more strongly
bound, and the variable-temperature data confirms the
increased binding enthalpies seen from the H2 adsorption data.
Electronic Structure Calculations. Density functional
theory (DFT) was used to examine the differences in the
electronic structures at the open metal sites by using
representative complexes to model the M2(dobdc) and
M2(m-dobdc) frameworks. DFT has shown itself to be a fairly
robust method for modeling chemical systems; its major
failings, self-interaction error and lack of dispersion inter-
actions,34 are well-known and can be corrected for by using
appropriate functionals that account for these failings, namely
range separation and explicit dispersion correction.35 Pre-
viously, the ωB97X-D36 functional was shown to accurately
model H2 binding in metal−organic frameworks.37
Modeling M2(dobdc)-type frameworks accurately with
electronic structure theory is challenging because fragmenting
the structure at any point will lead to the neglect of important
interactions from the chains formed by the M2+ ions. In an
attempt to understand the nature of the differences in hydrogen
binding between M2(dobdc) and M2(m-dobdc), some accuracy
in the calculated energetics is sacrificed in order to learn
relevant information about the differences between the two
systems. In this vein, our modeling focused on the linker of
interest coordinated to two Co2+ centers; the remaining ligands
on each Co2+ ion were truncated as formaldehyde molecules in
order to maintain charge balance. Since the key open
coordination site for each Co2+ center in the framework is
enforced by constraints on the ligands imposed by the
macrostructure of the framework, which could not be explicitly
included in our model, an unconstrained optimized geometry
of the model would not actually reflect the structural properties
of the framework in question. Thus, the geometry of the linker
complex was taken from the experimentally determined crystal
structure of the framework and subsequently frozen before the
interaction with H2 was optimized. Furthermore, the Co···H2
distance was taken from the neutron diffraction spectra and
used as a constraint to ensure an accurate depiction of this
interaction in the model system.
While comparing absolute energy differences gives a single
number to describe H2 binding, energy decomposition analysis
(EDA) breaks down that number into physically interpretable
components. An EDA based on absolutely localized molecular
orbitals (ALMOs)21 breaks down total binding energies into
frozen energy, polarization, and charge-transfer components.
The frozen term is due to permanent electrostatics and Pauli
repulsions, since the H2 electron density is being brought
within the van der Waals radius of the Co atom. The
polarization term corresponds to the favorable interaction of
electrons in the H2 and complex fragments relaxing in the
presence of the other fragment without electron transfer.
Charge transfer stems from energy lowering when electrons are
allowed to flow from one fragment to the other. A recent
generalization of the ALMO EDA to open-shell molecules
allows for the application of this method to general metal-
containing systems.38
Table 4. ALMO Energy Decomposition Analysis of H2
Binding to the dobdc4− and m-dobdc4− Complexes
energy (kJ/mol)
component
dobdc4−
m-dobdc4−
frozen
polarization
charge transfer
total
1.3
7.9
−4.8
−12.5
−16.0
−9.5
−17.1
−18.6
repulsion that stems from a shorter Co···H2 distance in the m-
dobdc4− complex. The increased charge-transfer energy of H2
binding in this species is key, as that is the largest term in the
total calculated binding energy. A breakdown of the charge-
transfer term can be seen in Table 5; in both cases, the largest
Table 5. Forward and Backbonding Contributions to the
Charge-Transfer Term for the dobdc4− and m-dobdc4−
Complexes
energy (kJ/mol)
dobdc4−
m-dobdc4−
H2 → Co bonding
−9.1
−3.4
−11.8
−5.3
Co → H2 backbonding
portion of the charge transfer comes from the H2 σ orbital to
the unoccupied orbitals of the Co complex. The increased H2
to Co forward donation is indicative of more positive charge at
the metal cation in the m-dobdc4− complex as compared with
the dobdc4− complex. This explains the experimentally
determined increase in binding enthalpy of H2 in the M2(m-
dobdc) frameworks as compared with their M2(dobdc)
analogues, as increased positive charge at the metal center
leads to stronger polarization of the bound H2. Additionally, a
difference is seen in the location of the bound H2 relative to the
linker in the two complexes. In the dobdc4− linker complex, the
H2 binds to create a nearly octahedral geometry around the
Co2+ ion, with a distance of 3.30 Å from the H2 to the aromatic
carbon bonded to the carboxylate group (the α carbon).
Conversely, in the m-dobdc4− complex, the H2 is reoriented
toward the linker at a distance of 2.64 Å from the α carbon
(Figure 8a,b). This difference suggests that a change in the
charge at the α carbon due to the different symmetry of the m-
dobdc4− linker possibly facilitates a change in the binding
geometry for H2. While this significant change may be an
artifact associated with the small cluster models, it is the same
trend seen experimentally for the H2···α carbon distance, which
is 3.48(3) Å in Co2(m-dobdc) and 3.59(2) Å in Co2(dobdc).
The orbital interactions between H2 and the Co complexes
provide further insight into the stronger H2 binding in Co2(m-
dobdc). While increased forward bonding is a product of
increased charge density at the metal, the nature of the
backbonding from the complex to H2 is quite different. This
can be seen by looking at the complementary occupied virtual
orbital pairs (COVPs) that contribute most to backbonding
(Figure 8c,d). In the dobdc4− complex, the occupied orbital is
primarily localized on the metal center, while in the m-dobdc4−
species, the donating orbital includes contributions from the π
system of the linker. This increased backbonding for m-dobdc4−
agrees with the redshift seen in the H2 infrared spectra for the
M2(m-dobdc) frameworks, as increased backbonding will
weaken the H−H bond. This possible extra interaction with
the linker in the m-dobdc4− complex, coupled with the larger
The binding energy decomposition analysis clarifies the
differences in these two species (Table 4). Specifically, there is
more charge transfer and polarization in the m-dobdc4−
containing complex, which is accompanied by a partially
offsetting increase in the energetically unfavorable frozen term.
This larger frozen term is the result of the increased steric
12126
dx.doi.org/10.1021/ja506230r | J. Am. Chem. Soc. 2014, 136, 12119−12129

Journal of the American Chemical Society
Article
neutron diffraction atomic parameters, Rietveld refinement
plots, additional VTIR spectra, COVPs, and additional
experimental details. This material is available free of charge
via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
jrlong@berkeley.edu
■
Present Address
¶Department of Chemistry and Biochemistry, Mount Allison
University, Sackville, New Brunswick, Canada E4L 1G8
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Research at Berkeley and NIST was supported through the
Department of Energy, Office of Energy Efficiency and
Renewable Energy (DOE-EERE), Fuel Cell Technologies
Office and through the Department of Energy, Office of
Science Graduate Fellowship Program (DOE SCGF), made
possible in part by the American Recovery and Reinvestment
Act of 2009, administered by ORISE-ORAU under contract no.
DE-AC05-06OR23100. Efforts at Oberlin College were funded
through NSF Grant CHE-1111896. We thank the 11-BM staff
at the Advanced Photon Source at Argonne National
Laboratory for assisting with powder X-ray diffraction experi-
ments. Use of the Advanced Photon Source at Argonne
National Laboratory was supported by the U.S. Department of
Energy, Office of Science, Office of Basic Energy Sciences,
under contract no. DE-AC02-06CH11357. We thank the NSF
for providing graduate fellowship support for J.A.M. and
M.T.K., the Camille and Henry Dreyfus Postdoctoral Program
in Environmental Chemistry and the NSERC of Canada for
fellowship support of S.J.G., and the NIST National Research
Council for a Postdoctoral Fellowship award to M.R.H. In
addition, we are grateful to Dr. Greg Halder for help with the
diffraction data collection on 17-BM-B at the APS, Dr. Trudy
Bolin for the use of the helium glovebox in APS Sector 9, Paul
Horn for assistance with the energy decomposition analysis,
and Eric Bloch and Rebecca Triano for helpful discussions.
Figure 8. Calculated binding modes of H2 in (a) the dobdc complex
and (b) the m-dobdc complex, in both of which the Co2+ is purple and
the H2 is above the Co2+. The corresponding COVPs are shown for
(c) the dobdc complex and (d) the m-dobdc complex, with occupied
and virtual orbitals shown as solid and mesh, respectively, at a 0.05
Å−3/2 isodensity.
forward donation of the H2 to the more positively charged
metal center, are the key differences between these two systems
that help to explain the stronger H2 binding seen in M2(m-
dobdc) versus M2(dobdc).
CONCLUSIONS
■
The foregoing results demonstrate the synthesis of a new family
of metal−organic frameworks, M2(m-dobdc) (M = Mg, Mn, Fe,
Co, Ni), representing a potentially less expensive version of
their well-known structural isomers M2(dobdc). Hydrogen was
used as an initial probe to determine how electronic and
structural differences in the new frameworks modified the
interactions of the open metal coordination sites with
adsorbates. The Mn, Fe, Co, and Ni variants exhibited greater
isosteric heats of H2 adsorption by approximately 1 kJ/mol
compared to the M2(dobdc) compounds. The results from
powder neutron diffraction, inelastic neutron scattering, and
infrared spectroscopy experiments performed on hydrogen-
loaded samples of Co2(m-dobdc) all support this as arising
from stronger interactions between the metal centers and the
H2 molecules. Future efforts will focus on the larger-scale
production of M2(m-dobdc) compounds and determining
whether the enhanced charge density at the metal sites leads
to improvements compared to their M2(dobdc) analogues in
the efficacy of key gas separations, including O2 from air,1j,39
CO2/N2,1h,j,7a,40 CO2/H2,1j,41 CO/H2,1e,39a,42 ethylene/etha-
ne,1i,9 propylene/propane,1i,3d,9 and methane purification.43
REFERENCES
■
(1) (a) Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.;
O’Keeffe, M.; Yaghi, O. M. Science 2002, 295, 469. (b) Kitagawa, S.;
Kitaura, R.; Noro, S.-I. Angew. Chem., Int. Ed. 2004, 43, 2334.
(c) Millward, A. R.; Yaghi, O. M. J. Am. Chem. Soc. 2005, 127, 17998.
(d) Ma, S.; Sun, D.; Simmons, J. M.; Collier, C. D.; Yuan, D.; Zhou,
H.-C. J. Am. Chem. Soc. 2008, 130, 1012. (e) Britt, D.;
Tranchemontagne, D.; Yaghi, O. M. Proc. Natl. Acad. Sci. U.S.A.
2008, 105, 11623. (f) Li, J.-R.; Kuppler, R. J.; Zhou, H.-C. Chem. Soc.
Rev. 2009, 38, 1477. (g) Britt, D.; Furukawa, H.; Wang, B.; Glover, T.
G.; Yaghi, O. M. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 20637.
(h) Sumida, K.; Rogow, D. L.; Mason, J. M.; McDonald, T. M.; Bloch,
E. D.; Herm, Z. R.; Bae, T.-H.; Long, J. R. Chem. Rev. 2012, 112, 724.
(i) Bloch, E. D.; Queen, W. L.; Krishna, K.; Zadrozny, J. M.; Brown, C.
M.; Long, J. R. Science 2012, 335, 1606. (j) Li, J.-R.; Sculley, J.; Zhou,
H.-C. Chem. Rev. 2012, 112, 869.
(2) (a) Suh, M. P.; Park, H. J.; Prasad, T. K.; Lim, D.-W. Chem. Rev.
2012, 112, 782. (b) Konstas, K.; Osl, T.; Yang, Y.; Batten, M.; Burke,
N.; Hill, A. J.; Hill, M. R. J. Mater. Chem. 2012, 22, 16698.
(3) (a) Dinca, M.; Dailly, A.; Liu, Y.; Brown, C. M.; Neumann, D. A.;
̆
Long, J. R. J. Am. Chem. Soc. 2006, 128, 16876. (b) Dietzel, P. D. C.;
ASSOCIATED CONTENT
* Supporting Information
Additional framework characterization, powder diffraction
patterns, thermogravimetric analyses, infrared spectra, BET
isotherm fits, adsorption isotherms, details on isotherm fitting,
■
S
Johnsen, R. E.; Blom, R.; Fjellvag, H. Chem.Eur. J. 2008, 14, 2389.
̊
(c) Vitillo, J. G.; Regli, L.; Chavan, S.; Ricchiardi, G.; Spoto, G.;
12127
dx.doi.org/10.1021/ja506230r | J. Am. Chem. Soc. 2014, 136, 12119−12129

Journal of the American Chemical Society
Article
Dietzel, P. D. C.; Bordiga, S.; Zecchina, A. J. Am. Chem. Soc. 2008, 130,
8386. (d) Bae, Y.-S.; Lee, C. Y.; Kim, K. C.; Farha, O. K.; Nicklas, P.;
Hupp, J. T.; Nguyen, S.; Snurr, R. Q. Angew. Chem., Int. Ed. 2012, 51,
1857. (e) Li, Y.-W.; Li, J.-R.; Wang, L.-F.; Zhou, B.-Y.; Chen, Q.; Bu,
Z.-H. J. Mater. Chem. A 2013, 1, 495. (f) Duan, X.; Cai, J.; Yu, J.; Wu,
C.; Cui, Y.; Yang, Y.; Qian, G. Microporous Mesoporous Mater. 2013,
181, 99. (g) Cai, J.; Yu, J.; Xu, H.; He, Y.; Duan, X.; Cui, Y.; Wu, C.;
Chen, B.; Qian, G. Cryst. Growth Des. 2013, 13, 2094.
(4) (a) Zhou, W.; Wu, H.; Yildirim, T. J. Am. Chem. Soc. 2008, 130,
15268. (b) Wu, H.; Zhou, W.; Yildirim, R. J. Am. Chem. Soc. 2009, 131,
4995. (c) Queen, W. L.; Bloch, E. D.; Brown, C. M.; Hudson, M. R.;
Mason, J. A.; Murray, L. J.; Ramirez-Cuesta, A. J.; Peterson, V. K.;
Long, J. R. Dalton Trans. 2012, 41, 4180.
Doerksen, R. J.; Dreuw, A.; Dunietz, B. D.; Dutoi, A. D.; Furlani, T. R.;
Gwaltney, S. R.; Heyden, A.; Hirata, S.; Hsu, C.-P.; Kedziora, G.;
Khalliulin, R. Z.; Klunzinger, P.; Lee, A. M.; Lee, M. S.; Liang, W. Z.;
Lotan, I.; Nair, N.; Peters, B.; Proynov, E. I.; Pieniazek, P. A.; Rhee, Y.
M.; Ritchie, J.; Rosta, E.; Sherrill, C. D.; Simmonett, A. C.; Subotnik, J.
E.; Woodcock, H. L., III; Zhang, W.; Bell, A. T.; Chakraborty, A. K.;
Chipman, D. M.; Keil, F. J.; Warshel, A.; Hehre, W. J.; Schaefer, H. F.,
III; Kong, J.; Krylov, A. I.; Gill, P. M. W.; Head-Gordon, M. Phys.
Chem. Chem. Phys. 2006, 8, 3172.
(19) (a) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem.
Phys. 1980, 72, 650. (b) McLean, A. D.; Chandler, G. S. J. Chem. Phys.
1980, 72, 5639.
(20) Dolg, M.; Wedig, U.; Stoll, H.; Preuss, H. J. Chem. Phys. 1987,
(5) Rosi, N. L.; Kim, J.; Eddaoudi, M.; Chen, B.; O’Keeffe, M.; Yaghi,
86, 2123.
O. M. J. Am. Chem. Soc. 2005, 127, 1504.
(6) (a) Dietzel, P. D. C.; Panella, B.; Hirscher, M.; Blom, R.; Fjellvag,
H. Chem. Commun. 2006, 959. (b) Liu, Y.; Kabbour, H.; Brown, C. M.;
Neumann, D. A.; Ahn, C. C. Langmuir 2008, 24, 4772. (c) Sumida, K.;
Brown, C. M.; Herm, Z. R.; Chavan, S.; Bordiga, S.; Long, J. R. Chem.
Commun. 2011, 47, 1157.
(21) Khaliullin, R. Z.; Cobar, E. A.; Lochan, R. C.; Bell, A. T.; Head-
Gordon, M. J. Phys. Chem. A 2007, 111, 8753.
(22) Khaliullin, R. Z.; Bell, A. T.; Head-Gordon, M. J. Chem. Phys.
̊
2008, 128, 184112.
(23) Duren, T.; Millange, F.; Fer
Phys. Chem. C 2007, 111, 15350.
́
ey, G.; Walton, K. S.; Snurr, R. Q. J.
̈
(7) (a) Dietzel, P. D. C.; Besikiotis, V.; Blom, R. J. Mater. Chem.
2009, 19, 7362. (b) Makal, T. A.; Li, J.-R.; Lu, W.; Zhou, H.-C. Chem.
Soc. Rev. 2012, 41, 7761. (c) Peng, Y.; Krungleviciute, V.; Eryazici, I.;
Hupp, J. T.; Farha, O. K.; Yildirim, T. J. Am. Chem. Soc. 2013, 135,
11887.
(24) (a) Dietzel, P. D. C.; Georgiev, P. A.; Eckert, J.; Blom, R.;
Strassle, T.; Unruh, T. Chem. Commun. 2010, 46, 4962. (b) Mason, J.
̈
A.; Veenstra, M.; Long, J. R. Chem. Sci. 2014, 5, 32.
(25) (a) Bhatia, S. K.; Myers, A. L. Langmuir 2006, 22, 1688.
́
(b) Garrone, E.; Bonelli, B.; Otero Arean, C. Chem. Phys. Lett. 2008,
456, 68. (c) Bae, Y.-S.; Snurr, R. Q. Microporous Mesoporous Mater.
2010, 132, 300.
(26) Irving, H.; Williams, J. P. Nature 1948, 162, 746.
(27) Brown, C. M.; Ramirez-Cuesta, A. J.; Her, J.-H.; Wheatley, P. S.;
Morris, R. E. Chem. Phys. 2013, 427, 3.
(8) (a) Dietzel, P. D. C.; Johnsen, R. E.; Fjellvag, H.; Bordiga, S.;
̊
Groppo, E.; Chavan, S.; Blom, R. Chem. Commun. 2008, 5125.
(b) Caskey, S. R.; Wong-Foy, A. G.; Matzger, A. J. J. Am. Chem. Soc.
2008, 130, 10870. (c) McDonald, T. M.; Lee, W. R.; Mason, J. A.;
Wiers, B. M.; Hong, C. S.; Long, J. R. J. Am. Chem. Soc. 2012, 134,
7056. (d) Yu, D.; Yazaydin, A. O.; Lane, J. R.; Dietzel, P. D. C.; Snurr,
R. Q. Chem. Sci. 2013, 3544.
(28) FitzGerald, S. A.; Pierce, C. J.; Rowsell, J. L. C.; Bloch, E. D.;
Mason, J. A. J. Am. Chem. Soc. 2013, 135, 9458.
(29) Kong, L.; Roman
Rev. Lett. 2009, 103, 096103.
́ ́
-Perez, G.; Soler, J. M.; Langreth, D. C. Phys.
(9) (a) Bao, Z.; Alnemrat, S.; Yu, L.; Vasiliev, I.; Ren, Q.; Lu, X.;
Deng, S. Langmuir 2011, 27, 13554. (b) He, Y.; Krishna, R.; Chen, B.
Energy Environ. Sci. 2012, 5, 9107. (c) He, Y.; Zhou, W.; Krishna, R.;
Chen, B. Chem. Commun. 2012, 48, 11813. (d) Wu, H.; Gong, Q.;
Olson, D. H.; Li, J. Chem. Rev. 2012, 112, 836. (e) Geier, S. J.; Mason,
J. A.; Bloch, E. D.; Queen, W. L.; Hudson, M. R.; Brown, C. M.; Long,
J. R. Chem. Sci. 2013, 4, 2054. (f) Herm, Z. R.; Bloch, E. D.; Long, J. R.
Chem. Mater. 2014, 26, 323.
(30) Forster, P. M.; Eckert, J.; Heiken, B. D.; Parise, J. B.; Yoon, J.
W.; Jhung, S. H.; Chang, J.-S.; Cheetham, A. K. J. Am. Chem. Soc. 2006,
128, 16846.
(31) FitzGerald, S. A.; Burkholder, B.; Friedman, M.; Hopkins, J. B.;
Pierce, C. J.; Schloss, J. M.; Thompson, B.; Rowsell, J. L. C. J. Am.
Chem. Soc. 2011, 133, 20310.
(32) (a) Garrone, E.; Bonelly, B.; Otero Arean
2008, 456, 68. (b) Otero Arean, C.; Chavan, S.; Cabello, C. P.;
Garrone, E.; Palomino, G. T. ChemPhysChem 2010, 11, 3237.
(33) Arean, C. O.; Chavan, S.; Cabello, C. P.; Garrone, E.; Palomino,
́
, C. Chem. Phys. Lett.
(10) (a) Liu, J.; Wang, Y.; Benin, A. I.; Jakubczak, P.; Willis, R. R.;
LeVan, M. D. Langmuir 2010, 26, 14301. (b) Liu, J.; Benin, A. I.;
Furtado, A. M. B.; Jakubczak, P.; Willis, R. R.; LeVan, M. D. Langmuir
2011, 27, 11451.
́
́
G. T. ChemPhysChem 2010, 11, 3237.
(11) Makal, T. A.; Yakovenko, A. A.; Zhou, H.-C. J. Phys. Chem. Lett.
(34) (a) Perdew, J. P.; Zunger, A. Phys. Rev. B 1981, 23, 5048.
2011, 2, 1682.
́
(b) Kristyan, S.; Pulay, P. Chem. Phys. Lett. 1994, 229, 175. (c) Dutoi,
(12) Sun, L.; Miyaki, T.; Seki, S.; Dinca,
135, 8185.
̆
M. J. Am. Chem. Soc. 2013,
A. D.; Head-Gordon, M. Chem. Phys. Lett. 2006, 422, 230.
(35) (a) Wu, X.; Vargas, M. C.; Nayak, S.; Lotrich, V.; Scoles, G. J.
Chem. Phys. 2001, 115, 8748. (b) Wu, Q.; Yang, W. J. Chem. Phys.
2002, 116, 515. (c) Chai, J.-D.; Head-Gordon, M. J. Chem. Phys. 2008,
128, 084106.
(13) (a) Schlapbach, L.; Zuetl, A. Nature 2001, 414, 353. (b) Collins,
̈
D. J.; Zhou, H.-C. J. Mater. Chem. 2007, 17, 3154. (c) van den Berg, A.
W. C.; Arean
́
, C. O. Chem. Commun. 2008, 668. (d) Murray, L. J.;
Dinca, M.; Long, J. R. Chem. Soc. Rev. 2009, 38, 1294.
̆
(36) Chai, J.-D.; Head-Gordon, M. Phys. Chem. Chem. Phys. 2008, 10,
6615.
(14) Lei, G.-D.; Li, L.-X.; Yu, Z. Y.; Xie, M. G. Chin. Chem. Lett.
2005, 16, 1039.
(37) Sumida, K.; Stuck, D.; Mino, L.; Chai, J.-D.; Bloch, E.;
(15) Udovic, T. J.; Brown, C. M.; Leao, J. B.; Brand, P. C.; Jiggets, R.
D.; Zeitoun, R.; Pierce, T. A.; Peral, I.; Copley, J. R. D.; Huang, Q.;
Neumann, D. A.; Fields, R. J. Nucl. Inst. Methods Phys. Res., Sect. A
2008, 588, 406.
(16) Azuah, R. T.; Kneller, L. R.; Qiu, Y.; Tregenna-Piggott, P. L. W.;
Brown, C. M.; Copley, J. R. D.; Dimeo, R. M. J. Res. Natl. Inst. Stand.
Technol. 2009, 114, 241.
(17) FitzGerald, S. A.; Churchill, H. O. H.; Korngut, P. M.; Simmons,
C. B.; Strangas, Y. E. Rev. Sci. Instrum. 2006, 77, 093110.
(18) Shao, Y.; Fusti Molnar, L.; Jung, Y.; Kussmann, J.; Ochsenfeld,
C.; Brown, S. T.; Gilbert, A. T. B.; Slipchenko, L. V.; Levchenko, S. V.;
O’Neill, D. P.; DiStasio, R. A., Jr.; Lochan, R. C.; Wang, T.; Beran, G. J.
O.; Besley, N. A.; Herbert, J. M.; Lin, C. Y.; Van Voorhis, T.; Chien, S.
H.; Sodt, A.; Steele, R. P.; Rassolov, V. A.; Maslen, P. E.; Korambath,
P. P.; Adamson, R. D.; Austin, B.; Baker, J.; Byrd, E. F. C.; Dachsel, H.;
̈
Zavorotynska, O.; Murray, L.; Dinca,
̆
M.; Chavan, S.; Bordiga, S.;
Head-Gordon, M.; Long, J. R. J. Am. Chem. Soc. 2013, 135, 1083.
(38) Horn, P. R.; Sundstrom, E. J.; Baker, T. A.; Head-Gordon, M. J.
Chem. Phys. 2013, 138, 134119.
(39) (a) Kerry, F. G. Industrial Gas Handbook: Gas Separation and
Purification; CRC Press: Boca Raton, FL, 2007. (b) Bloch, E. D.;
Murray, L. J.; Queen, W. L.; Chavan, S.; Maximoff, S. N.; Bigi, J. P.;
Krishna, R.; Peterson, V. K.; Grandjean, F.; Long, G. J.; Smit, B.;
Bordiga, S.; Brown, C. M.; Long, J. R. J. Am. Chem. Soc. 2011, 133,
14814.
(40) (a) Yazaydin, A. O.; Benin, A. I.; Faheem, S. A.; Jakubczak, P.;
Low, J. J.; Willis, R. R.; Snurr, R. Q. Chem. Mater. 2009, 21, 1425.
́
(b) Ferey, G.; Serre, C.; Devic, T.; Maurin, G.; Jobic, H.; Llewellyn, P.
L.; De Weireld, G.; Vimont, A.; Daturi, M.; Chang, J.-S. Chem. Soc. Rev.
12128
dx.doi.org/10.1021/ja506230r | J. Am. Chem. Soc. 2014, 136, 12119−12129

Journal of the American Chemical Society
Article
2011, 40, 550. (c) Li, J.-R.; Ma, Y.; McCarthy, M. C.; Sculley, J.; Yu, J.;
Jeong, H.-K.; Balbuena, P. B.; Zhou, H.-C. Coord. Chem. Rev. 2011,
255, 1791.
(41) Herm, Z. R.; Swisher, J. A.; Smit, B.; Krishna, R.; Long, J. R. J.
Am. Chem. Soc. 2011, 133, 5664.
(42) (a) McCandless, F. P. Ind. Eng. Chem. Process Des. Dev. 1972, 11,
470. (b) Miyajima, H.; Kodama, A.; Goto, M.; Hirose, T. Adsorption
2005, 11, 625. (c) Karra, J. S.; Walton, K. S. Langmuir 2008, 24, 8620.
(43) (a) Zhang, Y. F.; Musseman, I. H.; Ferraris, J. P.; Balkus, K. J. J.
́
Membr. Sci. 2008, 313, 170. (b) Bastin, L.; Barcia, P. S.; Hurtado, E. J.;
Silva, J. A. C.; Rodrigues, A. E.; Chen, B. J. Phys. Chem. C 2008, 112,
1575. (c) Bae, Y. S.; Farha, O. K.; Spokoyny, A. M.; Mirkin, C. A.;
Hupp, J. T.; Snurr, R. Q. Chem. Commun. 2008, 4135. (d) Mu, B.; Li,
F.; Walton, K. S. Chem. Commun. 2009, 2493. (e) Lee, J. Y.; Roberts, J.
M.; Farha, O. K.; Sarjeant, A. A.; Scheidt, K. A.; Hupp, J. T. Inorg.
Chem. 2009, 48, 9971. (f) Couck, S.; Denayer, J. F. M.; Baron, G. V.;
́
Remy, T.; Gascon, J.; Kaptejin, F. J. Am. Chem. Soc. 2009, 131, 6326.
(g) Lee, K.; Isley, W. C., III; Dzubak, A. L.; Verma, P.; Stoneburner, S.
J.; Lin, L.-C.; Howe, J. D.; Bloch, E. D.; Reed, D. A.; Hudson, M. R.;
Brown, C. M.; Long, J. R.; Neaton, J. B.; Smit, B.; Cramer, C. J.;
Truhlar, D. G.; Gagliardi, L. J. Am. Chem. Soc. 2014, 136, 698.
12129
dx.doi.org/10.1021/ja506230r | J. Am. Chem. Soc. 2014, 136, 12119−12129